"JBI will do for integration and the ESB what EJBs did for application logic and the application server," said David Chappell, vice president and chief technology evangelist for Sonic Software, the inventor and leading provider of the enterprise service bus (ESB), and member of the JSR 208 expert group. - http://xml.coverpages.org/JBI-EarlyReview.html

Interesting comment on member of JSR 208. What does it actually mean? Is JSR 208 going to do all the same mistakes that EJB did? Does it mean that JSR will "simplify" integration projects like EJBs did? Today, there are "zero code" integration tools available. How much simpler can it get?

Or maybe the JBI component model allows you to change container, in which your component reside. How many of you have actually changed your EJB container? I have not.

What are your thoughts on JSR 208, Java business integration?

Comment made by David Chappell got me thinking. Do we actually need another component model?
One goal of JSR is to reduce integration costs. That's simply great. There is no point of paying 100K$ for TIBCO
just to do integration within small organization. "Zero code" integration. Great!. But it still costs 5 times more
than doing actual integration with some proprietary solution using JMS.

In my opinion we need lightweight framework aka "Spring for Integration" that can be used for simple integration projects. There is another point. Are there "simple integration projects"?

I believe that there would be need for simple, cheap, BPEL based open source solution. If someone shares these thoughts, please let me know.

I don't really understand the motivation behind JBI either. The request doc reads like so much gobbledegook.
Can anyone elucidate?